i Developmental cancers represent a unique and heterogeneous spectrum of neoplasia that provide an excellent model system for defining clinically and therapeutically relevant categories of diseases and the identification of therapeutic targets. Present approaches for risk assessment and therapeutic assignment rely primarily on clinical findings in conjunction with select biological features. There is currently no means to determine the most appropriate therapeutic program based on the biology of a tumor. High throughput gene expression analysis allows rapid assessment of thousands of molecular events reflecting tumor physiology and can be used to address these critical issues. We have performed a comprehensive expression analysis of 100 neuroblastic tumors and in conjunction with Dr. Gorlick (Preclinical Core) and Dr. Ladanyi (Project 2) have expression data on more than 100 additional developmental solid tumors. This large data set provides an unprecedented opportunity to explore the biology of developmental solid tumors and the potential of genome wide characterization for classification and therapeutic target discovery. Our overall research goal is to use these large molecular data sets of well- characterized clinical tumor samples to identify critical molecules and pathways that define therapeutically relevant molecular subclasses of developmental tumors and provide therapeutic opportunities. This effort incorporates direct interactions with all other projects and cores in this program and provides a bridge between basic and clinical investigations. Our specific aims are: 1) To define and validate a molecular classification for clinically relevant subclasses of developmental tumors by; a) Development and testing ofbiostatistical/bioinformatic classification algorithms based on gene expression; b) Validation of the final classification models in independent tumor data sets through cooperative efforts and prospective studies; c) Characterization of select differentially expressed genes that strongly correlate with clinical subgroups and likely contribute to distinct tumor biology by confirming differential expression, and analysis of the expression and function of the corresponding protein. 2) To identify and characterize expressed genes and pathways that provide potential targets for drug and immune based therapy within appropriate disease states and test targeted agents by: a) discovery of targets by mathematical and bioinformatic approaches based on logical algorithms that include frequency of target presence in clinical states and normal tissues, predicted molecular function within active biological pathways and subcellular location; b) In vitro and in vivo testing of available targeted therapeutic agents identified in CMAP and Ingenuity databases, published literature and efforts in this program.